This invention relates to analogues of human parathyroid hormone, which have been found to be effective in the treatment of osteoporosis.
Osteoporosis is a leading cause of disability in the elderly, particularly elderly women. It has recently been realized that human parathyroid hormone (hPTH) and certain analogues are stimulators of bone growth that are useful in the treatment of osteoporosis. Osteoporosis is a progressive disease which results in the reduction of total bone mass and increased bone fragility. This often results in spontaneous fractures of load-bearing bones and the physical and mental deterioration characteristic of immobilizing injuries. Postmenopausal osteoporosis is caused by the disappearance of estrogens which trigger a decade-long acceleration of bone turnover with an increased imbalance between resorption of old bone and formation of new bone. This results in thinning, increased porosity, and trabecular depletion of load-bearing bones. Osteoporosis is also associated with hyperthyroidism, hyperparathyroidism, Cushing""s syndrome, and the use of certain steroidal drugs. Remedies historically have involved increase in dietary calcium, estrogen therapy, and increased doses of vitamin D, but mainly with agents such as antiresorptives that inhibit bone resportion by osteoclasts.
Parathyroid hormone (PTH) is produced by the parathyroid gland and is a major regulator of blood calcium levels. PTH is a polypeptide and synthetic polypeptides may be prepared by the method disclosed by Erickson and Merrifield, The Proteins, Neurath et al, Eds., Academic Press, New York, 1976, page 257, and as modified by the method of Hodges et al (1988), Peptide Research 1, 19, or by Atherton, E. and Sheppard, R. C., Solid Phase Peptide Synthesis, IRL Press, Oxford, 1989.
When serum calcium is reduced to below a normal level, the parathyroid gland releases PTH and the calcium level is increased by resorption of bone calcium, by increased absorption of calcium from the intestine, and by increased renal reabsorption of calcium from nascent urine in the kidney tubules. Although continuously infused low levels of PTH can remove calcium from the bone, the same low doses, when intermittently injected can actually promote bone growth.
Tregear, U.S. Pat. No. 4,086,196, described human PTH analogues and claimed that the first 27 to 34 amino acids are the most effective in terms of the stimulation of adenylyl cyclase in an in vitro cell assay. Rosenblatt, U.S. Pat. No. 4,771,124, disclosed the property of hPTH analogues wherein Trp23 is substituted by amino acids phenylalanine, leucine, norleucine, valine, tyrosine, xcex2-naphthylalanine, or xcex1-naphthylalanine as a PTH antagonist. These modified hPTH analogues also have the 2 and 6 amino terminal acids removed, resulting in loss of most agonist activities when used to treat osteoporosis. These analogues were designed as inhibitors or PTH and PTH-related peptide. The analogues were claimed as possibly useful in the treatment of hypercalcemia associated with some tumors,
Pang et al, WO93/06845, published Apr. 15, 1993, described analogues of hPTH which involve substitutions of Arg25, Lys26, Lys27 with numerous amino acids, including alanine, asparagine, aspartic acid, cysteine, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine, or valine. These are claimed, with no supporting data from animal or human trials, to be effective in the treatment of osteoporosis with minimal effects on blood pressure and smooth muscle.
PTH operates through activation of two second messenger systems, Gs-protein activated adenylyl cyclase (AC) and Gq-protein activated phospholipase Cxcex2. The latter results in a stimulation of membrane-bound protein kinase Cs (PKC) activity. The PKC activity has been shown to require PTH residues 29 to 32 (Jouishomme et al (1994) J. Bone Mineral Res. 9, (1179-1189). It has been established that the increase in bone growth, i.e. that effect which is useful in the treatment of osteoporosis, is coupled to the ability of the peptide sequence to increase AC activity. The native PTH sequence has been shown to have all of these activities. The hPTH-(1-34) sequence is typically shown as (A):
Ser Val Ser Glu lie Gln Leu Met His Asn Leu Gly Lys His Leu Asn Ser Met Glu Arg Val Glu Trp Leu Arg Lys Lys LeuGln Asp Val His Asn Phe-OH (SEQ ID NO:22)A
The following linear analogue, hPTH-(1-31)-NH2, for which data is included in Table 1, below, has only AC-stimulating activity and has been shown to be fully active in the restoration of bone loss in the ovariectomized rat model (Rixon, R. H. et al (1994) J. Bone Miner. Res. 9, 1179-1189; Whitfield et al (1996), Calcified Tissue Int. 58, 81-87; Willick et al, U.S. Pat. No. 5,556,940 issued Sep. 17, 1996):
Ser Val Ser Glu Ile Gln Leu Met His Asn Leu Gly Lys His Leu Asn Ser Met Glu Arg Val Glu Trp Leu Arg Lys Lys Leu Gln Asp Val-NH2B
The above molecule, B SEQ ID NO:32, and its counterpart with a Leu27 substitution SEQ ID NO:2 may have a free carboxyl ending instead of the amide ending illustrated.
It is an object of the present invention to produce new PTH analogues with greater metabolic stability, increased bone restoration activity, increased AC activity, and minimal clinical side effects.
According to one aspect of the invention, human parathyroid hormone hPTH and pharmaceutically acceptable salts thereof are provided, having the amino acid sequence
R-NH-R1-Val-Ser-Glu-Ile-Gln-Leu-R2-His-Asn-Leu-Gly-Lys-R3-R4-R5-R6-R7-Glu-Arg-Val-R8--Trp-Leu-R9--R10--R11-Leu-R12-Asp--Y (SEQ ID NO:23)
wherein,
R=hydrogen or any linear or branched chain alkyl, acyl or aryl group,
R1=Ser, Ala or Aib,
R2=Met, or a naturally occurring hydrophobic amino acid,
R3=His or a water soluble amino acid,
R4=Leu or a water soluble amino acid,
R5=Asn or a water soluble amino acid,
R6=Ser or a water soluble amino acid,
R7=Met, or a naturally occurring hydrophobic amino acid,
R8=Glu, Lys or Asp,
R9=Arg, Cys, Lys, Orn, or His
R10=Arg, Lys, Orn, Gln, Glu or Asp
R11=a naturally occurring hydrophobic or polar amino acid,
R12=Gln, Arg, Glu, Asp, Lys or Orn,
Cterm=OH, NH2, and
Y=Cterm, Val-Cterm, Val-His-Cterm, Val-His-Asn-Cterm, Val-His-Asn-Phe-Cterm, Val-His-Asn-Phe-Val-Cterm, Val-His-Asn-Phe, Val-Ala-Cterm and Val-His-Asn-Phe,Val-Ala-Leu-Cterm, (SEQ ID NOS:24-27)
cyclized as between one or two amino acid pairs 22 and 26, 26 and 30, 27 and 30, and 25 and 29 are lactams when R9 is Lys or Orn and R12 is Glu or Asp, excluding cyclo(Lys26-Asp30)(Leu27)-hPTH-(1-34)-NH2, cyclo(Lys27-Asp30)-h-PTH-(1-34)-NH2 and cyclo(Lys26-Asp30)-(Leu27)-hPTH-(1-34)-OH.
Examples of the salts include salts of inorganic acids, salts of organic acids such as formic acid, acetic acid, tartaric acid and citric acid, salts of inorganic bases such as sodium and ammonium and salts of organic bases such as triethylamine, ethylamine and methylamine.
According to another feature of the present invention, cyclisation is effected by the formation of lactams, involving the coupling of the side-chains of the selected amino acid pairs such as between natural residues 22 and 26, or 26 and 30. Other types of cyclisations, such as the formation of a disulfide bridge e.g., between Cys containing analogues Cys22-Cys26 and Cys26-Cys30 are also contemplated.
Substitutions of various amino acids have also been found to be effective. Lys27 may be replaced by a Leu or by various other naturally occurring hydrophobic or polar residues. Another factor is how well the residue fits to the receptor. Ala is not as hydrophobic as Leu. Lys and Tyr are generally considered to be polar, but nonetheless have hydrophobic interactions with the receptor. Lys, for example, can fold so that the hydrophobic part interacts with other hydrophobic residues in the receptor, and the NH2 is exposed to solvent. Such substitutions include ornithine, citrulline, xcex1-aminobutyric acid, alanine, norleucine, isoleucine and tyrosine, or any linear or branched xcex1-amino aliphatic acid, having 2-10 carbons in the side chain, any such analogue having a polar or charged group at the terminus of the aliphatic chain. Examples of polar or charged groups include: amino, carboxyl, acetamido, guanido and ureido. Although it appears that Leu27 is the best substitution, it also appears that many other pos27 substitutions retain nearly full activity and could also have desired properties, such as increased proteolytic stability or water solubility. Ile, norleucine, Met, and ornithine are expected to be the most active.
This substitution results in a stabilization of an xcex1-helix in the receptor-binding region of the hormone. This has been confirmed by examination of the circular dichroism spectrum of the lactam analogues, as compared to the circular dichroism spectrum of the linear molecule, (Leu27)-hPTH-(1-31)-NH2. Circular dichroism spectra are highly sensitive to the presence of xcex1-helical secondary structure, and the technique has been used to demonstrate the presence of xcex1-helix in hPTH fragments (Neugebauer et al (1991) Biochemistry31, 2056-2063). Furthermore, the stabilization of xcex1-helix on formation of the above-mentioned lactams in hPTH-(20-34)-NH2 has been shown (Neugebauer et al (1994) Int. J. Protein Peptide Res. 43, 555-562). There is a potential amphiphilic xcex1-helix between residues 21 and 31 of hPTH-(1-31)-NH2, and data has been presented showing that the hydrophobic face of this helix interacts with the PTH receptor (Neugebauer, W. (1995) et al Biochemistry 34, 8835-8842; Gardella, T. J. et al (1993), Endocrinology 132, 2024-2030).
It has been found that the most effective cyclisation involves the formation of a lactam, for example, between either residues Glu22 and Lys26, or Lys26 and Asp30. Other cyclisations are also possible such as between Lys27 and Asp30, although this lactam has been found to exhibit some de-stabilizing effect on the xcex1-helix.
More specifically, receptor-binding studies of PTH fragments have indicated a principal binding region within residues 14-34.1 We have suggested that the residues 17-29 xcex1-helix binds as such to the PTH receptor, and that the amphiphilic portion of this xcex1-helix binds with its hydrophobic face to the receptor.2 This model is consistent with the results of a study of receptor binding-region analogues.
NMR studies have shown that even a model peptide found to be the structure of a receptor-bound peptide hormone, such as PTH, cannot be inferred reliably from its free structure in solution. Constrained analogues of peptide hormones have been used to limit the number of conformational states available to the peptide8. Examination of the sequence of hPTH reveals 3 possible salt bridges within residues 17-29 which could either stabilize or destablize xcex1-helix. These are between Glu22 and Lys26, and Lys26 and Asp30, both of which are expected to stabilize an xcex1-helix, and between Lys27 and Asp30, which is expected to destablize an xcex1-helix.4 Lactam formation between these residue pairs would restrict the conformations available to hPTH in this helical region. Furthermore, two of these luctams, Glu22-Lys26 and Lys26-Asp30 which are expected to stabilize xcex1-helical structure are located on the polar face of the amphiphilic portion of the xcex1-helix. The third one, Lys2714 Asp30, is expected to at least partially destabilize xcex1-helix and involes a residue, Lys27, which is on the hydrophobic face of the amphiphilic helix. Cyclisation as between positions 25 and 29 can occur if Lys or Orn replaces Arg in position 25, and if Gln29 is replaced with Glu or Asp.
The substitution of Leu for the Lys27 results in a more hydrophobic residue on the hydrophobic face of the amphiphilic helix. This resulted in increased adenylyl cyclase stimulating activity in the ROS cell time. It will be appreciated by those skilled in the art that other such substitutions discussed above would likely result in analogues with the same or increased activities.
The combined effect of substitution and either lactam formation is expected to stabilize the xcex1-helix and increase bioactivity, and to protect this region of the molecule from proteolytic degradation. The presence of the amide at the C-terminus is preferred in the sense that it is further expected to protect the peptide against exoproteolytic degradation, although some peptidases can hydrolyze them. (Leslie, F. M. and Goldstein, A. (1982) Neuropeptides 2, 185-196).
It has also been found that other amino acid substitutions can usefully be made. Specifically, we have replaced the oxidation sensitive Met residue at positions 8,18 with a naturally occurring hydrophobic residue, Nle, as per Japanese Patent publication 61-24598. It is also to be expected that other such hydrophobic residues like Leu, Ile, Val, Phe and Trp would also be useful, as per U.S. Pat. No. 5,393,869 to Nakagawa et al.
Reverse lactams are also contemplated. For example, we have shown the effectiveness of a Lys22-Glu26 switch. It is therefore to be expected that similar switches could usefully be made as between the 26-30 and 27-30 lactams.
Another substitution at the preferred 22-26 lactam site, in addition to the aforementioned Cys-Cys, ie Asp22-Orn26 has been done to illustrate that different cyclisation/ring sizes can usefully be made.
In the U.S. Pat. No. 5,393,869 Nakagawa et al and U.S. Pat. No. 5,434,246, Fukuda et al some substituted hPTH analogues were reported to have substantial AC activity and might have enhanced stabilities to proteolytic attack, specifically.
1. Ser-1 to Aib (xcex1-aminolsobutyric acid)
2. Lys-27 to Gln (reported to have 2.5xc3x97AC activity)
3. Residues 14, 15, 16, 17 to Lys, in whole or in part (reported to greatly increase activityxe2x80x94up to 8xc3x97. This may be due to increase in water solubility. In vivo, these are expected to be more labile to trypsin-like enzymes). They claim this tetrapeptide (residues 14-17, incl.) such that there is at least one water soluble amino acid. For example His-14 or Lys-14; Leu-15, Lys-15 or Arg-15; Asn-16, Orn-16, Hci-16, Asp-16, Arg-16, Lys-16, D-Lys-16, Ser-16 or Gly-16; and Ser-17, Lys-17, Asp-17 or Arg-17. Could also include Glu, for example.
4. Arg 25 to His to minimize protease attack. Since our lactams, particularly with Leu or another hydrophobic amino acid at position-27, can become somewhat insoluble and also difficult to dissolve, it would be expected that the same substitutions would be useful in our lactams.
It will also be appreciated by those skilled in the art that although the 1-31 h-PTH cyclics may be preferred, it is to be expected from the data presented herein that cyclic fragments in the range of 1-30 to 1-37 will also be effective. In particular, there is no evidence in the literature that the presence of additional amino acids up to 37 affect the biological properties of the hormone, particularly given the confirmatory 1-34 data included herein.
The lactams according to the invention may be prepared by known procedures described below, and may be used for stimulating bone growth, for restoring bone, and for the promotion of bone healing in various circumstances, such as in the treatment of osteoporosis and normal fractures.